Communications systems, such as Voice over Internet Protocol (VoIP) networks, real-time streaming multimedia networks, and other real-time IP communications networks, have become widely used for varying purposes ranging from transmitting media for entertainment to transmitting media for videoconferencing. In these systems, bandwidth resources are reserved based on statistical models that anticipate a given amount of concurrent streams, calls, or sessions. For example, the administrator or operator generally chooses a set of bandwidths to be used during calls, and each endpoint joining is automatically assigned a bandwidth for the duration of the call.
Such communications systems, however, often fail when the capacity needed to serve the customers that wish to use their subscribed services is not available. These communications systems attempt a number of approaches to ensure a particular quality of service (QoS). For example, when a customer on a VoIP communications system attempts to setup a new call and there is not enough bandwidth available, the VoIP communications system typically refuses to setup a call when the available bandwidth limit has been reached. Alternatively, some VoIP communications systems reduce the call quality of existing streams, calls, or sessions (e.g., using a different encoding technique) that are utilizing the available bandwidth resources, thereby freeing up enough bandwidth resources to allow the additional call to be placed. The mechanisms used to perform this functionality are raw and underdeveloped, and do not allow the service provider to provide an optimal service to its customers. As a result, these approaches often fail to meet the QoS guarantees given by service providers and often cause these service providers to reimburse their customers for failing to meet a particular level of service.
There is therefore a need in the art for approaches that determine and assign optimal media characteristics within a constrained communications resource. Accordingly, it is desirable to provide devices, methods, and media that overcome these and other deficiencies of the prior art.